New release-plan

[Aleix Pol]

On Fri, Nov 27, 2009 at 10:53 PM, David Nolden <zwabel at googlemail.com>wrote:

> Am Freitag 27 November 2009 22:32:45 schrieb Alexander Dymo:
> >
> > Which usually means _never_ release.
> >
> > I personally think that:
> >
> > 1) The direction Andreas took (to remove all unfinished/not working
> stuff)
> >  is the only right way to do the release
> >
> > 2) We need to release now. KDevelop is a big project, you will always
> want
> >  to "finish" things here and there. 1 month of delay won't change
> anything
> >  in this respect
> >
> > 3) We need to face the fact, 4.0 release will suck anyway. It won't be
> >  feature complete and it won't work well. Remember, 3.0 wasn't that good
> >  either. It's 3.1 release which was really good.
> But I won't participate in that suckage. Why do we need to do it? It gains
> us
> nothing. I hate that KDE4 has done it that way..
>
> I don't expect the release to be feature-complete, but a basic set of
> properly
> working features should be there, else we don't even need to bother
> releasing
> it. Why call something "4.0" if it really is a beta?
>
> > 4) To stabilize, let's continue to throw out and disable features. That's
> >  not as bad as you, guys think. Commercial software development has
> taught
> >  me that people can live without surprising amount of features, just as
> >  long as your app doesn't crash ;)
>
> It is a consensus that we don't want unstable features. If a non-critical
> feature is unstable and not well maintained, then it is fine to remove it.
> However I, and some others, consider the source-formatter stuff a critical
> feature, which should be fixed (in any way) instead of removed.
>
> Greetings, David
>
> --
> KDevelop-devel mailing listI agree that
> KDevelop-devel at kdevelop.org
> https://barney.cs.uni-potsdam.de/mailman/listinfo/kdevelop-devel
>

I, and I think most KDevelop team, think that fixing is better than
disabling. We are a small team and we have to stay together, we don't want
the group falling apart anyway, because we are KDevelop

The fact we should agree on is, anyway, that we need to release as soon as
we have anything decent to start to show it off to people. Dropping a
feature or waiting to fix it is a matter of taste. So as Andreas said, we
all agree it's ok to wait a little bit more if it's going to mean a better
release. Fine, let's wait a little longer.

On the git matter, it would be nice. I had this fantasy where KDE would move
itself and we wouldn't have to start moving project by project. We want to
stay with KDE on that, because it simplifies a lot the translation issues,
but I also understand that there are quite some benefits in git. Here we
could try to follow amarok's steps, they're using it and quite happy with
that. Also we would need someone to take care about that, it's quite some
work to move such a project (much more than a svn mv).

Let's be positive
Aleix
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kdevelop-devel/attachments/20091127/b00634e3/attachment.html>